Hook
Alert: The International Monetary Fund just dropped a working paper that redefines the stablecoin threat vector. Brandon Joel Tan’s March 2025 model isn’t about reserve adequacy or smart contract bugs. It’s a mathematical proof that stablecoins can transform from harmless liquidity tools into coordinated weapons of mass capital flight under fixed exchange rate regimes. I’ve watched the Terra death spiral from the trenches. This is worse. This is systemic, state-dependent, and regulators are already taking notes.
Context
Stablecoins today are a $170 billion market. USDT alone accounts for over $100 billion. They’re the backbone of DeFi, the on-ramp for millions in emerging markets, and the preferred refuge for anyone fleeing hyperinflation. But the IMF paper argues their role is not static. Under normal conditions, they enable efficient exchange, reduce transaction costs, and provide a dollar-denominated savings vehicle that outpaces local bank deposits. That’s the welfare gain. The catch: when a country’s currency becomes severely overvalued relative to its fixed peg, stablecoins become a coordination device. They allow every rational actor to simultaneously exit the local currency at minimal friction. The result is a self-fulfilling crisis that accelerates the very devaluation everyone fears.
Tan’s model is rooted in game theory. In a fixed-rate system with capital controls, agents face a collective action problem: they want to sell local currency for dollars, but doing so individually risks detection or penalties. Stablecoins solve that. By offering a pseudonymous, liquid, and globally accessible dollar asset, they collapse the coordination cost. The model shows a binary outcome: either the peg holds and stablecoins are benign, or the peg is perceived as weak and stablecoins become the primary channel for a run. The speed of that run is orders of magnitude faster than traditional bank runs because there’s no settlement lag, no bank holiday, no withdrawal limit.
Core
Let’s unpack the mechanics. Tan’s paper uses a formal model with two states: “normal” and “distress.” In the normal state, agents hold a mix of domestic currency and stablecoins. The domestic currency offers a liquidity premium; stablecoins offer stability. The welfare gain from stablecoins in this state is clear: they expand the choice set. In distress, defined as when the fixed exchange rate becomes unsustainable due to fundamental misalignment, the model introduces a threshold effect. Once a critical mass of agents switches into stablecoins, the remaining holders of domestic currency face a growing incentive to follow. This is the coordination mechanism. The paper provides a closed-form solution for the threshold, linking it to variables like the size of the stablecoin market, the cost of moving between assets, and the credibility of the central bank.
Empirical backing comes from real-world data. Argentina’s parallel market exchange rate has consistently diverged from the official rate by 40-100% since 2020. During that period, stablecoin volumes on local exchanges surged from under $10 million daily to over $200 million. Turkey saw a similar pattern in 2021-2022 when the lira collapsed. The paper quantifies the elasticity: a 1% increase in the black-market premium leads to a 0.8% increase in stablecoin trading volume in those markets. That’s not correlation; that’s causation through the mechanism Tan describes.
From my experience reverse-engineering the Terra collapse in 2022, I can confirm the coordination problem. UST’s peg broke not because of a fundamental imbalance alone, but because the arbitrage mechanism created a feedback loop. Every second the peg drifted, more holders had an incentive to sell, and the algorithm’s response (expanding LUNA supply) only amplified the exit. The IMF model formalizes that same dynamic for fiat-backed stablecoins in emerging markets. The difference is that here, the coordination is not between bots on a blockchain but between millions of retail users with phones. The speed is terrifying.
Contrarian
The paper is elegant, but it’s blind to two critical flaws. First, it treats stablecoins as monolithic. USDT, USDC, DAI, and BUSD have fundamentally different risk profiles. USDT’s reserves are opaque; USDC’s are audited. DAI is overcollateralized and decentralized. In a crisis, these differences matter. The model assumes a single stablecoin asset class with uniform liquidity and trust. In reality, during a coordinated run, the weakest stablecoin (likely USDT due to reserve concerns) would fail first, creating a flight-to-quality within stablecoins. That internal dynamic could actually dampen the systemic risk because capital would stay in stablecoins rather than fleeing to offshore dollars. The IMF’s model misses this internal flight channel.
Second, the paper ignores the issuer’s role as a potential circuit breaker. Tether has the ability to freeze addresses, comply with OFAC sanctions, and even pause redemptions under extreme conditions. In a crisis, issuers could act as a speed bump, buying time for central banks to respond. The paper treats stablecoins as a pure market mechanism, but they are centrally managed financial products. A disciplined issuer could, for example, impose a temporary fee on conversions or limit daily minting. That would contradict the narrative of stablecoins as autonomous accelerants. Surveillance isn’t just watching; it’s anticipating the break before it happens. The IMF fails to incorporate the human gatekeeper in its model.
I’ve seen this before. In 2017, I audited HotCo’s smart contract and found an integer overflow that would have let an attacker drain $2 million. The fix was trivial, but the vulnerability was invisible in normal usage. That’s the state-dependent risk the IMF paper highlights, but it’s also a warning: models that ignore operator intervention are dangerous. A stablecoin issuer with a kill switch is not an accelerant; it’s a fire extinguisher that might or might not work. The paper’s policy implication—that regulators should restrict stablecoin use in fixed-rate regimes—overlooks the possibility that well-designed stablecoins with strong issuer discipline are actually less risky than the informal dollarization they replace.
Takeaway
The IMF has given a mathematical foundation for tighter regulation. The real question is not whether stablecoins can amplify crises; they can. The question is whether the alternative—dollar shortages, black markets, and capital controls—is worse. Yield is the bait; liquidity is the trap. The next currency crisis will test whether regulators understand the model’s limits. Watch USDT’s reserves. Watch Argentina’s premium. The signal will be a sudden spike in USDT/ARS trading volume. When that happens, don’t panic. Anticipate the break. And remember: a red candle doesn’t lie.